Retinoblastoma has played a pivotal role in cancer research that far outweighs its epidemiologic significance. Despite its rarity, retinoblastoma has provided unique and far-reaching insights into the genetics and molecular biology of cancer. The autosomal dominant inheritance of retinoblastoma has attracted the attention of cancer researchers for many years. The pivotal breakthrough occurred in 1971, when Alfred Knudson proposed his two-hit hypothesis that proposed that retinoblastoma results from the mutational inactivation of both alleles of a putative recessive oncogene [13]. The retinoblastoma locus was mapped to chromosome 13q through its linkage to esterase D, an enzyme that had previously been mapped to this region and which could be assayed for its enzymatic activity [26]. Esterase D activity was found to be about half the normal level in retinoblastoma patients with karyotypically visible deletions on chromosome 13q [27]. In another study, esterase D activity was 50 % of normal in healthy tissues of retinoblastoma patients but was undetectable in tumor tissue [28], providing further support for Knudson’s hypothesis . The recessive nature of the retinoblastoma locus was conclusively demonstrated by loss of heterozygosity in tumor tissue [29], a technique that subsequently has become a standard for studying tumor suppressor genes. Based on the proximity to the esterase D gene, several groups independently identified and characterized a nearby gene that proved to be the retinoblastoma gene [22, 30, 31]. Reintroduction of this gene into retinoblastoma cells suppressed the tumor phenotype, providing functional evidence for its role as a tumor suppressor [32].

The retinoblastoma gene (RB) spans over 200 kilobases of DNA and contains 27 exons [33, 34]. Germ line mutations tend to cluster at CpG areas but are also spread throughout the gene with no mutational hot-spot [35, 36]. Germ line mutations can include deletions, frameshift and nonsense mutations, splicing mutations, and mutations in introns resulting in a truncated protein product [35]. The second hit is usually a nondisjunction or recombination event during mitosis, resulting in reduction to hemizygosity of the mutant allele [37]. Most RB mutations result in >90 % penetrance, but some mutations have been associated with low penetrance retinoblastoma. These low penetrance mutations can be divided into those that cause a reduction in the amount of normal Rb protein (usually promoter mutations) and those that result in a partially functional Rb protein (usually missense mutations) [38].

The retinoblastoma protein (Rb) is composed of 928 amino acids and has three functional domains: the N-terminus , the central pocket domain, and the C-terminus . The pocket, which is formed by the highly conserved A and B boxes, is required for binding to E2F, viral oncoproteins and chromatin remodeling proteins, and it is required for tumor suppression [39, 40]. Most RB mutations affect the pocket [35]. The C-terminus is also required for efficient binding to E2Fs, and it has binding sites for the oncoproteins MDM2 and c-abl [41–44]. The function of the N-terminus is less clear, and it does not appear to be required for tumor suppression [45].

A major function of Rb is to inhibit the G₁–S phase transition of the cell cycle [46]. Protein activity is modulated during the cell cycle by phosphorylation, which can occur on 16 cyclin-dependent kinase Ser/Thr-Pro phosphoacceptor sites. Rb phosphorylation varies throughout the cell cycle, with hypophosphorylated Rb predominating in G₀ and G₁, and hyperphosphorylated Rb in S-phase and G₂/M [47–49]. Rb usually becomes hypophosphorylated during differentiation and senescence, and it becomes hyperphosphorylated in cells reentering the cell cycle [50–52]. The Rb protein can also be modulated by viral oncoproteins, such as SV40 large T antigen, adenovirus E1a protein and herpes virus E7 protein, which bind and inhibit Rb, thereby stimulating cell cycle progression [53–55] (Fig. 23.4).

A major binding partner of Rb is the E2F family of transcription factors (E2Fs) [56]. Rb does not have a DNA binding domain but is brought to specific promoters through its interaction with E2Fs. E2F-responsive genes contain one or more E2F binding sites in their promoters, and these genes play an important role in cell cycle, apoptosis and other cellular functions [57]. In addition to a DNA binding domain, E2Fs 1–5 also have a transactivation domain by which they promote the transcription of specific genes. Rb inhibits transcription in part by binding and masking the E2F transactivation domain [58, 59]. In addition, Rb actively represses promoter and enhancer elements independently of its effect on the E2F transactivation domain [60, 61]. This so-called active repression is now thought to result from the recruitment by Rb of chromatin remodeling proteins that reconfigure local chromatin into a nonpermissive state for transcription [62]. These proteins include histone deacetylases, DNA methyltransferases, and polycomb complexes [63–68]. The functional significance of these two independent mechanisms of transcriptional repression by Rb remains controversial (Fig. 23.3). However, it seems clear that active repression by Rb is required for the full tumor suppressor activity of Rb, including its inhibition of the G₁–S phase transition [69] (Fig. 23.4).

In addition to its role as a cell cycle inhibitor, Rb also inhibits apoptosis [46]. These seemingly opposing roles of Rb have not been satisfactorily reconciled. The widely accepted view of Rb as an on-off regulator of the G1–S transition that is completely inactivated every cell cycle does not explain why cells do not undergo apoptosis with every attempt at cell division. The most straightforward explanation is that Rb is not completely inactivated every cell cycle but that the degree of Rb phosphorylation and inactivation can be graded, depending on the cellular context. In support for this idea, Rb is phosphorylated in several hierarchical steps, and the phosphorylation events regulating the cell cycle can be differentiated from those regulating apoptosis [70, 71]. It seems likely that Rb is only partially phosphorylated and inhibited during normal cell division, and that this incomplete inactivation is sufficient to block its cell cycle inhibitory function and to allow cell cycle progression without triggering apoptosis [72]. This residual anti-apoptotic activity may persist unless Rb is more completely phosphorylated (or unless the RB gene is mutated), suggesting that Rb may serve as a buffer against apoptosis during normal cell division and a checkpoint for triggering apoptosis under abnormal stress conditions such as excessive mitogenic stimulation, DNA damage, and hypoxia in which apoptosis may be favored for survival of the organism.

Although RB gene mutations are common in only a few types of cancer, such as retinoblastoma and small cell lung cancer, the RB pathway is functionally disrupted in virtually all human cancers [73, 74]. The RB pathway is composed of proteins that modulate Rb phosphorylation state. Rb is phosphorylated by cyclin-dependent kinases (CDKs) , which are activated by interaction with cyclins [74, 75]. Therefore, overactivity of cyclins or CDKs can lead to functional inactivation of Rb. For example, cyclin D1 is a bona fide oncogene that is overexpressed and functionally inhibits Rb in many cancer types [76, 77]. A constitutively active CDK4 mutant that maintains Rb in a phosphorylated state has been identified in some cancers [78]. On the other hand, CDKs are inhibited by proteins such as p16^INK4a, which is a tumor suppressor that is mutationally inactivated in some cancers [79]. Interestingly, the cancers associated with complete Rb inactivation by genetic deletion (e.g., retinoblastoma and small cell lung cancer) have a different phenotype than those associated with partial Rb inhibition by loss of p16Ink4a or overexpression of cyclin D (e.g., melanoma, most solid tumors). The former tend to exhibit rapid proliferation, high apoptotic rates and sensitivity to chemotherapy, whereas the latter tend to have slow rates of proliferation and apoptosis and tend to be resistant to chemotherapy [80].